Peer Revenue Fully funded

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Tuesday 15 September

High Energy Physics (Phenomenology)

Multiplicative NLO matching in DIS with Pythia 8

We present a new multiplicative next-to-leading order (NLO) matching scheme in Pythia 8 for deep-inelastic scattering (DIS) processes in electron-proton collisions. The implementation is based on numerical event-by-event integration over the virtual corrections and real-emission phase-space and matrix-element corrections (MECs) for the first parton-shower emission. The latter has been included for two parton-shower options currently available in Pythia 8 to handle DIS processes, the default shower with dipole-recoil option and Vincia. The implementation has been validated against other available NLO-capable event generators and compared to experimental data from the HERA collider. In all cases we find very good agreement confirming the validity of the new Pythia implementation. We also study the impact of the MECs in each shower and consider the implications for the upcoming Electron-Ion collider.
Read the full abstract

Networking

Beyond Measurement Metrics: A Human-Centered Framework for Semantic Validation of Network Traffic Classification

Machine learning (ML) has become the dominant approach for network traffic classification, achieving very high predictive performance. However, a model is only valuable if it learns semantically meaningful and trustworthy patterns rather than exploiting spurious correlations. Conventional evaluation practices predominantly assess predictive performance. Consequently, whether the model relies on semantically meaningful patterns remains unknown. To address these challenges, we adapt the knowledge generation framework for network traffic classification. The adapted framework combines data, ML models, explainability, visualization, and expert reasoning to support the iterative exploration, verification, and refinement of model behavior and data preprocessing. The framework is grounded in findings from the literature, benchmark dataset analyses, practical experience with XAI-based traffic classification, and expert feedback, providing practical guidance for semantic model validation. By complementing predictive performance with semantic validation and human expertise, the proposed framework supports the development of network traffic classification models that are not only accurate but also robust and trustworthy.
Read the full abstract

All authors Igor Cherepanov, David Sessler, Alex Ulmer, Thorsten May, Jörn Kohlhammer

Quantum Physics

Strictly Localized Mixed States

We study strictly localized mixed states in quantum field theory: states that are indistinguishable from vacuum outside some spacetime region. Such states are generated from the vacuum by so-called Licht maps. While the states associated with individual outcomes of local, projective measurements on the vacuum are not strictly localized, the corresponding non-selective states are strictly localized and even admit decompositions into strictly localized pure states. Remarkably, however, there are strictly localized mixed states that cannot be written as mixtures of pure states strictly localized to the same region. Such states can arise from unsharp local measurements. Finally, we extend Knight's theorem to mixed states, showing that strictly localized bosonic mixed states contain terms with arbitrarily high particle numbers.
Read the full abstract

Artificial Intelligence

ORDER: Task-Conditioned Routing for Retrieval-Augmented Generation

Retrieval-Augmented Generation (RAG) pipelines typically rely on a fixed indexing and retrieval configuration determined at preprocessing time. This one-size-fits-all design is ill-suited to domain-expert settings, where heterogeneous queries require different chunking granularities, metadata constraints, and source-selection strategies. As a result, configurations that are effective for one family of queries often perform poorly for others. In this paper, we introduce ORDER (Optimal Routing for Dynamic Evidence Retrieval), a query-conditioned RAG framework that jointly adapts indexing and retrieval to the incoming query. Our approach first discovers semantic clusters over a given set of questions associated to a corpus and learns, for each cluster, a chunking strategy together with a suited metadata filtering and reranking configuration. At inference time, queries are routed to the appropriate pre-built index through nearest-centroid assignment. To further improve retrieval, we propose a supervised query router (QRe) that predicts which collections are most likely to contain relevant evidence, coupled with a Uniform Multi-source Sampler (UMS) that allocates the retrieval budget evenly across the selected sources. We evaluate our framework on large-scale, heterogeneous historical archives and show that conditioning both indexing and retrieval on the query consistently outperforms both naive baselines and strong state-of-the-art RAG systems in complex expert-domain environments.
Read the full abstract

Systems and Control

On personal recommendations in social networks

Social networks in which algorithms actively influence humans through personal recommendations are ubiquitous. While opinion dynamics is an established tool to analyze these systems, existing models typically do not capture how individual agents process personal recommendations. In this work, we introduce a model for personal recommendations that is analytically tractable and consistent with the confirmation bias phenomenon from behavioral psychology. We describe how individuals process recommendations based on prior beliefs and a sensitivity parameter using a Gaussian influence function. Using this model, we analyze the effect of different recommendation policies. For broadcast policies, where recommendations are homogeneous across the agents, a bifurcation analysis shows that the system exhibits bistability, which may result in unintended consequences. For personally targeted policies, we first derive optimal personal recommendations and then extend to robust convergence when the agents' sensitivity to personal recommendations is uncertain. The collective behavior of the proposed model and the derived policies are illustrated and evaluated through numerical simulations.
Read the full abstract

Artificial Intelligence

ThinkFlow: Self-Evolving Probabilistic Latent Memory for Lifelong Conversational Agents

Lifelong conversational agents rely on memory systems to maintain deep, context-aware interactions with users. However, existing explicit textual memory pipelines suffer from a severe information bottleneck, often losing subtle behavioral patterns and emotional shifts. Furthermore, being typically static post-deployment, they cannot autonomously adapt to personal habits and preferences without manual feedback. Cognitive science, however, suggests that humans maintain mental models purely in a latent space and continuously refine them through predictive coding. Inspired by this, we propose ThinkFlow, a novel end-to-end latent memory framework for lifelong conversational agents. ThinkFlow bypasses the text bottleneck by dynamically compressing conversational flows into probabilistic latent memory skills, autonomously consolidating complex user states into disentangled, continuous vectors without semantic interference. To break this barrier, we introduce a test-time evolution paradigm. By coupling teacher-guided latent alignment to bootstrap the initial state with a self-supervised next-user-utterance prediction task for continuous refinement, the framework successfully overcomes cold-start challenges and achieves label-free lifelong personalization. Extensive experiments on long-term conversation benchmarks demonstrate that ThinkFlow significantly outperforms prevailing memory systems, providing highly personalized and contextually accurate responses over extended multi-session interactions.
Read the full abstract

All authors Cai Ke, Xin Liu, Han Zhang, Jiangyue Yan, Zike Yuan, Ling Deng, Yue Yu, Hui Wang, Ruifeng Xu

Quantum Physics

Indistinguishability of single Raman photons from single atoms

We theoretically investigate the indistinguishability of single photons generated from single trapped $^{40}$Ca$^+$ ions in a Raman scattering process driven by few-nanosecond excitation pulses. Of particular interest is how spontaneous decay back to the initial state affects Hong-Ou-Mandel (HOM) photon interference. Numerical simulations identify the mean number of back-decays as a measurable quantity that correlates with achievable HOM visibility. Optimization of the excitation pulse with respect to a trade-off between photon yield and indistinguishability is analyzed. Finally, we compare the performance of different trapped-ion species for long-range dual-rail entanglement swapping via HOM photon interference.
Read the full abstract

Artificial Intelligence

FlexEE: Self-Speculative and KV-Compatible Early Exiting for Offloading-Aware LLM Inference

Large language model (LLM) inference is often constrained by both computation and memory, especially in offloading-based deployments where model weights are transferred across memory hierarchies during autoregressive decoding. In this setting, reducing the number of executed layers can lower per-token latency while also avoiding costly weight movement. Motivated by this observation, we present FlexEE, an early exiting framework for resource-constrained and offloading-based LLM inference. FlexEE makes early exiting practical for LLM decoding through layer-wise exit supervision for reliable intermediate-layer prediction, self-speculative decoding over a Top-K local vocabulary for low-cost exit decisions, and dynamic hidden state management for KV-cache-correct and memory-aware execution. Across generative and downstream tasks, FlexEE enables efficient early exit with minimal accuracy degradation, delivering up to 1.27$\times$/3.16$\times$ and 1.25$\times$/2.83$\times$ end-to-end speedups on Llama2-7B and Llama3-8B under 0%/50% weight offloading, respectively.
Read the full abstract

Software Engineering

Search-Based Metamorphic Testing of Vision-Language Models in Autonomous Underwater Robotic Software

Our industry partner focuses on quality assurance for industrial systems across multiple domains, including maritime systems, such as overwater vessels and autonomous underwater robots (AURs). Despite the strong performance of vision-language models (VLMs) in scene understanding, image captioning, and object recognition, their use in AUR software operating in underwater environments is underexplored. Therefore, in this context, it is important to evaluate the quality of VLMs for integration into AUR software and, so, automated software testing tools are needed to assess their suitability and improve their dependability. To this end, we propose a search-based metamorphic testing approach (MetaVLM) that identifies a minimal set of transformations on underwater images to induce incorrect model predictions, thereby revealing VLM failures. We employ NSGA-II as a multi-objective search algorithm and evaluate it over open-source VLMs, BLIP and CLIP, against a random search baseline. Results demonstrate the strengths and limitations of each VLM in the context of AUR software systems. Based on the results, we derive lessons for software engineering practitioners and researchers working on quality assurance of VLM-based software systems.
Read the full abstract

All authors Muhammad Yousaf, Aitor Arrieta, Shaukat Ali, Paolo Arcaini, Shuai Wang

Group Theory

Unipotence of a double commutator with transvections

Let $R$ be a nonzero commutative ring with identity, $n\geq3$, and $\sigma\in GL_n(R)$. We prove that there exist nonidentity transvections $\tau_1,\tau_2\in GL_n(R)$ such that $\big[[\sigma,\tau_1],\tau_2\big]=I_n$. Hence, Kourovka Notebook Problem 10.46 has a stronger affirmative answer over commutative rings with identity. The proof uses rank-one operators and the existence of a nonzero functional annihilating a standard basis vector and its image under $\sigma$.
Read the full abstract

Soft Condensed Matter

Odd slip at chiral active surfaces

Odd slip at solid-liquid interfaces, characterized by a velocity component perpendicular to the applied shear stress, introduces a novel class of odd transport phenomena with transverse momentum transfer at boundaries, distinct from odd viscosity, odd elasticity, or odd diffusivity. This effect arises on chiral active surfaces with persistent energy input and defined handedness, such as carpets of rotating biological or synthetic cilia with flexible anchoring. Although odd slip itself is dissipationless, its maintenance in active systems comes at an energetic cost, which we estimate from the microscopic model. With respect to its orientation dependence, odd slip can be classified as pseudo-scalar or pseudo-vectorial. We use a perturbative expansion to determine the resistance tensors of problems with different levels of broken symmetries. They reveal how the surface chirality and body asymmetry together violate the Onsager symmetry, but uphold the Onsager-Casimir reciprocity. Strikingly, suspensions of particles with odd slip embedded in a conventional Newtonian bulk fluid exhibit odd viscosity, providing a new minimal mechanism for its emergence beyond current microscopic models.
Read the full abstract

Computer Vision

Symmetry-Aware Likelihood-Orbit Aggregation for Selective Left-Right Claim Verification

Frozen vision-language models (VLMs) remain unreliable on fine-grained left-right claims, and raw claim likelihoods need not reliably rank verification errors. After a horizontal-reflection intervention is fixed, how should its induced likelihood measurements be combined into a selective verification signal? We introduce Relation-Orbit, a closed-form contrast with no learned fusion parameters that assigns eight normalized likelihoods to query-supporting and counterfactual roles determined by reflection, inverse relation, and entity exchange. A claim is asserted only when the signed contrast exceeds a threshold selected on held-out data using pointwise Clopper-Pearson upper confidence bounds. On VSR and GQA across four frozen VLMs, Relation-Orbit yields higher mean test coverage at a 10% selective-risk calibration target than an all-eight Orbit-Max baseline in all eight dataset-backbone settings; gains over a nearly abstain-all one-sided intervention score are reported separately. A separate LLaVA-1.5/COCO evaluation, reduced-orbit controls, and a two-sided partition diagnostic further characterize the structural advantage.
Read the full abstract

All authors Zhouzhi Xiong, Chuxi Zhang, Weizhen He, Yi Chen, Qi Li, Donglian Qi

Number Theory

Refined conjectures on Fitting ideals of BDP Selmer groups

Under mild assumptions, we prove the strong main conjecture for BDP Selmer groups of elliptic curves with good reduction in the sense of Kurihara's refined Iwasawa theory. In particular, the structure of the initial Fitting ideal of BDP Selmer groups over finite subextensions in the anticyclotomic $\mathbb{Z}_p$-extension is determined by the values of the corresponding $p$-adic modular form at CM points.
Read the full abstract

Optimization and Control

From Consensus-Based Optimization to Particle Swarm Optimization: Convergence Guarantees under Drift-Diffusion Coupling

Particle swarm optimization (PSO) is a widely used algorithm featured in many state-of-the-art optimization tool-kits. However, rigorous performance guarantees are still lacking. The standard PSO dynamics do not admit a natural mean-field description, which would provide an avenue for theoretical analysis. By modifying the PSO formulation, one can recover the consensus-based optimization (CBO) algorithm with memory, which admits a mean-field limit and facilitates rigorous convergence analysis. These theoretical guarantees rely heavily on the fact that for CBO, the drift and noise strengths can be chosen independently, whereas they are coupled for PSO. We analyze how the PSO parameter coupling affects existing convergence guarantees for CBO and its variant with memory effect. We show, by an explicit construction, that the coupling still leaves a non-empty set of admissible parameters for these convergence guarantees to hold. However, the admissible parameter ranges shrink in the limits used to recover PSO. The resulting convergence guarantees from CBO therefore do not directly extend to the classical PSO model. We provide numerical simulations illustrating the parameter tradeoffs shown in the theoretical analysis.
Read the full abstract

Mesoscale and Nanoscale Physics

Nonlinear Hall Effect in Altermagnetic Warped Topological Insulators

Low-dimensional Dirac systems offer a versatile platform for quantum-geometric non-linear transport. Hexagonal warping in topological insulator (TI) surface states generates rich momentum-space geometric textures. However, mirror and time-reversal symmetries strictly constrain both the Berry curvature and quantum metric dipoles to zero. Here, we show that proximity-coupling a hexagonally warped TI to a d-wave altermagnet lifts spatial mirror symmetry via the interplay of C_{3v} warping and d-wave spin-splitting. This symmetry breaking enables finite Berry curvature dipole components, which, however, remain heavily suppressed near charge neutrality, establishing the intrinsic quantum metric dipole as the sole driver of the non-linear Hall response. This scattering-independent response vanishes at \mu = 0 and exhibits odd parity under chemical potential inversion. Importantly, rotating the altermagnetic orientation angle \phi continuously tunes quantum metric dipole and induces a sign reversal, peaking near \phi=0 and vanishing at \phi = \pi/4. Our findings highlight TI-altermagnet interfaces as ideal candidates for gate- and orientation-tunable quantum metric spintronics.
Read the full abstract

Statistics Applications

Combining Weather Forecast Aggregation and State-Space Models for Adaptive Probabilistic Electricity Load Forecasting

Accurate electricity load forecasting is essential to ensure the real-time balance between supply and demand, especially in systems increasingly influenced by weather conditions and renewable energy integration. In this paper, we propose an adaptive probabilistic forecasting framework that leverages multiple meteorological forecast providers to mitigate the impact of weather prediction errors on load forecasts. Our approach combines an adaptive generalized additive model (GAM) with a state-space formulation estimated via the Viking algorithm, enabling dynamic parameter updates over time. To improve uncertainty quantification, we extend the framework with a probabilistic layer based on GAM for location, scale and shape (GAMLSS), allowing the residual variance to depend on explanatory variables. We further introduce a short-term correction mechanism to enhance performance across different forecast horizons. A key contribution of this work is the aggregation of multiple meteorological inputs, whose diversity is shown to improve both point and probabilistic forecasts, particularly in capturing extreme quantiles by incorporating forecast dispersion. The methodology is evaluated on French national electricity consumption data and demonstrates consistent gains over single-provider approaches, with significant improvements in probabilistic forecasting accuracy.
Read the full abstract

Operator Algebras

Classifiable C*-algebras have quantum symmetries

We show that every classifiable simple C*-algebra, in particular the Jiang-Su algebra, admits outer actions of several discrete quantum groups, such as finite quantum groups and duals of certain free compact quantum groups. As a consequence of the construction, we provide new examples of hyperfinite subfactors with the Temperley-Lieb-Jones standard invariant and show that $\operatorname{SU}_q(2)$ can act minimally on some Araki-Woods factors for infinitely many $q$.
Read the full abstract

Analysis of PDEs

Large time nonlinear dynamics close to inhomogeneous stationary states of the Vlasov-HMF model

We consider the Vlasov-HMF (Hamiltonian Mean-Field) model and symmetric perturbations of symmetric inhomogeneous stationary states that are naturally associated with a pendulum Hamiltonian flow. We first show the existence of families of such stationary states having compact support inside the separatrix region, constant value near the origin, and satisfying a linear stability (Penrose) criterion. We then describe the nonlinear dynamics of small compactly supported perturbations with finite regularity of this class of equilibria. More precisely, we prove that for large (but finite) times, in a modulated action-angle coordinate system the solution remains close to a modulated stationary state and that the magnetization of the perturbation satisfes decay estimates driven by the linear dynamics.
Read the full abstract

Computation and Language

Can LLMs Follow the Pulse of a Crisis? Evaluating Crisis Sentiment in Bangladesh's July Uprising

Crisis sentiment analysis is especially challenging for low-resource languages such as Bangla, where language, context, and public reaction shift rapidly. We introduce UNRESTSENT200K, a Bangla crisis sentiment dataset with approximately 200K Facebook and YouTube comments from the July-August 2024 Bangladesh uprising. The dataset covers five event-aligned phases, from early escalation and internet blackout to regime transition and a later flood crisis. Each comment is linked to its parent post, enabling evaluation with and without discourse context. All comments are annotated through a fully human process involving 14 native Bangla-speaking annotators and senior validation, achieving substantial agreement (kappa = 0.73, alpha = 0.71) and 94.2% blind-audit agreement. We benchmark fine-tuned encoders, prompted LLMs, and LoRA-tuned LLMs. Results show that parent-post context consistently improves performance, while temporal shift across phases causes large performance drops. Strong LLMs perform well, but still struggle with sarcasm, implicit political references, and phase-dependent meaning. UNRESTSENT200K provides a benchmark for studying context-aware and temporally robust sentiment analysis in low-resource crisis discourse. UNRESTSENT200K is available at https://sami0055.github.io/UNRESTSENT200K/
Read the full abstract

All authors Md. Samiul Alim, Mahir Shahriar Tamim, Tanvir Ahmed Khan, Sharjil Khan, Rafia Ferdous Duti, Shahriyar Zaman Ridoy, Mohammad Ali Moni

Robotics

Overcoming technical adoption barriers for mobile service robots in rehabilitation

Many publications on robotic systems in healthcare describe early-stage work on low technology readiness levels. This paper describes how a mobile service robot approved as a medical device reaches higher technology readiness levels by adding peripheral functions and smaller improvements, which are pivotal for user acceptance in clinical environments and which often cannot be elicited by questioning users ex-ante as certain aspects only come in mind from testing the systems in clinical settings or operational environments. Especially developers of service robots in healthcare are advised to plan with such downstream developments, which can take significant implementation time, to obtain user acceptance and achieve widespread adoption of their robotic systems.
Read the full abstract

All authors Christian Sternitzke, Sebastian Blumenthal, Lukas Kleedoerfer, Verena Deserno, Anke Mayfarth

Computation and Language

PaperDoctor: Evidence-Grounded and Actionable Feedback for Scientific Papers in Progress

Autoresearch agents are reshaping the research ecosystem, but they can also let flawed claims enter the literature at scale. Human advisors catch such issues in drafts through careful, traceable feedback, yet advisor-style assessment requires extensive manual effort and does not scale. To shift automated paper assessment from a judge to a diagnostician, we introduce PaperDoctor, an agent framework for pre-submission feedback with three key innovations. First, a holistic hierarchical framework evaluates writing, layout, references, code, theory, prior work, and experiments through three layers: L1 surface screening, L2 typed verifiers that route each claim to the appropriate evidence, and L3 reproducers that rerun experiments by priority. Second, each finding contains an observation, a pointer to specific evidence such as a sentence, equation, or code line, and a revision suggestion, making critiques auditable and actionable. Third, PaperDoctor selectively rebuilds and reruns experiments based on claim importance and compute budget, surfacing reproducibility gaps and quantitative limitations that are invisible from the manuscript alone. We evaluate PaperDoctor on 30 in-progress papers, yielding 70.6% agreement and all positive holistic scores, and on 40 manuscripts across machine learning, natural science, and social science, covering human- and AI-authored papers with code. Overall, PaperDoctor produces more auditable feedback than human and other agentic reviewers, pairs critiques with concrete suggestions by design, and complements dimensions often overlooked by human reviewers. We also develop an interactive interface that lets authors browse findings grounded in their paper. PaperDoctor reframes automated paper assessment as diagnosis rather than verdict, taking a concrete step toward AI advisors for more rigorous AI-assisted scientific discovery.
Read the full abstract

All authors Kevin Qinghong Lin, Siyuan Hu, Pan Lu, Yu Chen, Yanzhe Chen, Owen Queen, Yupeng Chen, Jialin Yu, Junchi Yu, Zifeng Ding, Yuanfeng Ji, Sheng Liu, Jindong Gu, Linjie Li, Mike Zheng Shou, Philip Torr, James Zou

Computation and Language

The Role of Implicit and Explicit Demographic Signals in Large Language Model-based Student Assessment

Large Language Models are now common in student assessment, but we know little about how student demographics affect their use. Sometimes, considering student demographics may be necessary – for example, to improve readability for users with lower educational levels. However, it also risks being a cause of discrimination, e.g., when assigning lower scores to students from lower socioeconomic backgrounds. We set up controlled prompts to test 1) explicit demographic effects, where we mention demographic details directly, and 2) implicit effects, where we use conversation history as a demographic signal. We test these settings in three tasks: Automated Essay Scoring, Formative Feedback, and Metalinguistic Question Answering. We test six state-of-the-art LLMs on these tasks. In both explicit and implicit cases, the models pick up on demographic cues and can change their scoring, feedback, and answers accordingly. We find that LLMs frequently adjust the readability of feedback to education levels when these are explicitly mentioned. On the other hand, implicit conditions produce unpredictable biases, such as in question answering, where responses from lower-education levels receive lower sentiment scores. Our results provide clear evidence of demographic sensitivity in LLMs for educational assessment tasks.
Read the full abstract

Astrophysics of Galaxies

Lyman-$\alpha$ forest 1D flux power spectrum constraints on QSO-assisted reionization models

Recent JWST observations have revealed a large population of faint Quasi-Stellar Objects (QSOs) at redshifts 4 < z < 6, including a highly reddened sub-population of Little Red Dots. We exploit the sensitivity of the Lyman-$\alpha$ forest 1D flux power spectrum to thermal fluctuations in the Intergalactic Medium (IGM) to place statistical constraints on the spectral properties of the QSOs. By post-processing simulated Lyman-$\alpha$ forest spectra from the Sherwood-Relics suite of Lyman-$\alpha$ forest simulations with added HeII reionization by QSOs, in conjunction with published precision measurements of the Lyman-$\alpha$ forest 1D flux power spectrum between 4.2 < z < 5.0, we find that the addition of temperature boosts of the IGM within the HeIII regions improves agreement with the measured power spectra. While the contribution of the Little Red Dot population to HeII reionization is consistent with both a mild temperature boost of $\Delta T_b=1\times10^4$ K for soft spectra QSOs and $\Delta T_b=2\times10^4$ K for hard spectra QSOs, a contribution from the larger population of faint QSOs found by JWST having $M_{\rm UV} > -21.6$ is excluded at the 2$\sigma$ level if their spectra are sufficiently hard to boost the IGM temperature in HeIII regions by $\Delta T_b=2\times10^4$ K or greater.
Read the full abstract

Computation and Language

Autoformalizing Argumentative Material Inferences

Natural language arguments are compelling before they are formally explicit. A premise supports a claim through defeasible warrants, background commitments, and exception conditions that the text leaves implicit. However, formal verification requires the opposite. Making such arguments machine-checkable requires constructing the missing commitments, not only translating given sentences into logic. Construction, however, carries a risk that translation does not: a system free to add premises can make any claim provable, and a formally valid proof may assert the claim outright, prove it without the original premise, or establish more than the claim itself. We address this problem by formulating autoformalization for argumentative material inference as guard completion, in which non-monotonic material support is turned into monotonic formal inference relative to an explicitly constructed guard set. A completion is accepted only when its proof both passes the theorem prover and survives contrastive tests of premise dependence and claim selectivity. We implement this formulation in GUARD, a neuro-symbolic framework in which LLMs construct and formalize candidate guards, Isabelle/HOL verifies the resulting theories and returns step-level feedback for iterative refinement, and the system abstains when no faithful completion can be reached. Our empirical results on Debatepedia and ARCT using different LLMs demonstrate that GUARD yields significant improvements in verified-faithful (+35.3, +32.9 points) and substantial reductions in leakage (-25.9, -21.9 points) over the state-of-the-art LLM-driven theorem proving approach. Moreover, we show that the symbolic soft critique and the explicit assumption layer account for most of these gains, with the soft critique also improving the initial validity of the elicited context and reducing the number of iterations required for successful verification.
Read the full abstract

General Relativity and Quantum Cosmology

Impact of the Host Galaxy on the Search for Gravitational-Wave Microlensing by Compact Objects

Gravitational-wave (GW) microlensing occurs when an intervening compact object causes diffraction of the GW, imprinting characteristic frequency-dependent modulations on the observed signal. Current searches for microlensing signatures in the detected GW events, including those used to constrain the fraction of dark matter in the form of compact objects, assume an isolated point mass lens (PML) model. However, astrophysical compact objects are typically embedded within larger structures, such as galaxies, whose potentials introduce additional convergence and tidal shear, thereby perturbing the lensed signal. In this work, we study the impact of the background galaxy in the search for microlensing by compact objects using a PML model. We distribute the compact objects inside a galaxy following the Navarro-Frenk-White density profile. We model the microlensing effects using the potential of a PML embedded in a constant convergence and shear, which are, in turn, calculated from a singular isothermal sphere model of the galaxy lens. We demonstrate that neglecting the galaxy's effects can systematically reduce sensitivity in microlensing searches, particularly for high microlens masses. Thus, future searches may need to incorporate the additional effect of the lens galaxy, particularly those aimed at placing robust constraints on the abundance of compact dark matter.
Read the full abstract

All authors Uddeepta Deka, Apratim Ganguly, Soham Bhattacharyya, Parameswaran Ajith

Sound

Taming Long-form Text-to-Speech

Long-form text-to-speech (TTS) enables multi-turn conversations with consistent prosody and higher quality voice cloning from longer reference audio. Recent open-weights autoregressive TTS models such as Qwen3-TTS and VoxCPM2 attain state-of-the-art word error rate (WER) and speaker similarity (SIM) on short-form prompts but significantly deteriorate when used with long-form prompts. We propose Localized Attention-Constrained Inference (LACI), an inference-only method to detect TTS errors in near real-time, roll back to the error onset and regenerate with temporary guardrails, adding negligible computational overhead. Using LACI, we improve worst-of-N WER across 10 RNG seeds for Qwen3-TTS-0.6B from 35.2% to 3.4% on prompts longer than 1500 words, even surpassing its short-form reliability of 5.4% on prompts with fewer than 500 words. To demonstrate the efficacy of LACI on voice cloning reliability, we propose a sliding-window version of the SIM metric that we call wSIM. wSIM exposes several novel failure patterns that are not captured by SIM. LACI improves worst-of-N wSIM from 0.01 to 0.47 on 120 seconds of reference audio while reducing the rate of catastrophic generations with WER above 30% from 26% to below 1%
Read the full abstract

All authors Rongxiang Wang, Berkin Durmus, Aysegul Orhon, Eduardo Pacheco, Atila Orhon

Functional Analysis

Inner multipliers, similarity and invariant subspaces for a class of Harmonically weighted Dirichlet spaces

We consider the harmonically weighted Dirichlet spaces $D(\mu)$ induced by Borel measures which are mutually absolutely continuous with respect to the Lebesgue measure $m$ on the unit circle $\mathbb T$. Let $H^2$ and $D$ denote the Hardy space and the Dirichlet space on the unit disc $\mathbb D$, respectively. For any $f\in H^2$, let $m_f$ denote the measure defined by $dm_f(\zeta)=|f(\zeta)|^2 dm(\zeta)$. Our first result provides a characterization of functions in $D(m_f)$, when $f$ satisfies the condition $|f'(z)|^2=O(\frac{1}{(1-|z|)^{1-\epsilon}})$ for some $\epsilon>0$. We then study the multiplier algebras of these spaces with particular emphasis on inner multipliers. Specifically, when $f'$ is bounded, we obtain an explicit description of all singular inner functions in the multiplier algebra of $D(m_f)$ in terms of the radial zero set of $f$. As an application, we prove that if $f',g'$ are bounded and the radial zero sets of $f$ and $g$ are different, then the operators $(M_z, D(m_f))$ and $(M_z, D(m_g))$ are not similar. This generalizes a result of Richter showing that the operators $(M_z, D(m_{z-1}))$ and $(M_z, D)$ are not similar. Then, we classify all invariant subspaces $\mathcal M$ of $(M_z, D(\mu))$ for which $M_z|_{\mathcal M}$ is similar to $(M_z, D(\mu))$, where $\mu$ is any measure mutually absolutely continuous with respect to $m$. Finally, we study a special case of the problem of finding all functions $g\in D$ for which $gD(m_g)$ is a closed invariant subspace of $(M_z, D)$. A key ingredient in many of our results is the Richter-Sundberg formula for the local Dirichlet integral.
Read the full abstract

Software Engineering

TasmScan: Continuation-Aware Taint Analysis for TVM Bytecode with Savelist Abstraction

The Open Network (TON), with a peak market capitalization exceeding $20 billion and over 175 million activated on-chain addresses, relies on the TVM (TON Virtual Machine) to execute smart contracts. TVM uses first-class continuations with savelists to manage control flow and register state across continuation invocations. Since savelist-captured registers allow data to flow across continuation boundaries without passing through the operand stack, bytecode-level analyses cannot construct complete data flow tracking without explicitly modeling savelist semantics. We present TasmScan, the first bytecode-level static analysis framework for TVM that enables cross-continuation data flow reasoning without requiring source code. TasmScan models savelist semantics via forward register analysis with a formal over-approximation guarantee for exact-resolved save sites and locally tracked register definitions, then lifts bytecode into TASIR, a typed intermediate representation, and performs path-sensitive taint analysis with context-aware sources to detect defects. We evaluate TasmScan on 2,921 contracts from the TON verifier registry and a labeled benchmark of 208 contracts with human-confirmed ground truth. On the full corpus, TasmScan resolves 294,546 dynamic continuation targets with 100% precision; ablation confirms that savelist propagation is essential for resolving indirect register calls that depend on cross-continuation register passing. On the benchmark, TasmScan detects 95.3% of defects across five classes with 96.8% precision. A 366-pair stratified sample from the full corpus estimates 85.8% overall precision. TasmScan offers a 17x median speedup over the state-of-the-art symbolic-execution baseline, and in the path-analysis comparison completes 100% of analyses with zero crashes or timeouts.
Read the full abstract

Multiagent Systems

ToMAS: A Pilot Failure-Grounded Theory-of-Mind Benchmark from Multi-Agent LLM Failures

LLM-based multi-agent systems can fail even when communication succeeds because agents do not correctly track their peers' roles, knowledge, or intentions. We investigate whether such inter-agent misalignment cases, labelled FC2 in MAST-Data, can be converted into functional partner-state reasoning items. ToMAS applies four explicit convertibility criteria to diagnosed execution traces. A full conversion pass over 242 eligible non-AG2 training traces produced 39 CLEAN items. In an 18-trace reliability pilot, two annotators achieved 94.4% raw agreement and Cohen's kappa = 0.92. We then used the converted items as binary rewards in a small-scale GRPO feasibility experiment with Qwen2.5-1.5B. On a 28-item held-out Magentic GAIA diagnostic, every evaluated condition exceeded the ROUGE-L threshold on the same 2 of 28 items. Post-hoc adapter checks show why: under the learning rate used, the LoRA update remained numerically negligible (max abs Delta W about 7e-6), so all conditions decode identically to the untrained checkpoint. The experiment therefore does not show a training effect and cannot establish one; it reports an executable pipeline together with two limitations that any conclusive study must address: a provenance gap between the training and evaluation items, and lexical-overlap scoring. ToMAS provides a preliminary rubric and pipeline for converting diagnosed coordination failures into trainable partner-state reasoning items and identifies the requirements for a conclusive matched-domain evaluation.
Read the full abstract